[erlang-questions] Future of epmd

Kenneth Lundin kenneth.lundin@REDACTED
Thu Nov 8 17:55:32 CET 2012


Den 8 nov 2012 09:59 skrev "Kukosa, Tomas" <
tomas.kukosa@REDACTED>:
>
> Hi,
>
>
>
> BTW what about using SCTP for distribution protocol?
>
> Would not bring it some advatages in high availability area? E.g. bacause
of multihoming or posibility of setting timeout and retransmission
parameters?

Yes it would be interesting to have distribution over SCTP, and it is
possible to implement with the same plugin approach as the distro over SSL
is implemented.

This might have implications on epmd as well and maybe heterogenous
distribution
would be of interest as well.

With heterogenous distribution I mean that a node can talk sctp with some
other node and talk tcp with yet another. It would require some negotiation
and or registration in an extended epmd where a node can say which
protocols it supports and prefers.

/Kenneth, Erlang/OTP Ericsson
>
>
> Best regards,
>
>   Tomas
>
>
>
> From: erlang-questions-bounces@REDACTED [mailto:
erlang-questions-bounces@REDACTED] On Behalf Of Dmitry Demeshchuk
> Sent: Wednesday, November 07, 2012 8:03 AM
>
> To: erlang-questions
> Subject: [erlang-questions] Future of epmd
>
>
>
> Hello, list.
>
>
>
> As you may know, epmd may sometimes misbehave. Loses nodes and doesn't
add them back, for example (unless you do some magic, like this:
http://sidentdv.livejournal.com/769.html ).
>
>
>
> A while ago, Peter Lemenkov got a wonderful idea that epmd may be
actually written in Erlang instead. EPMD protocol is very simple, and it's
much easier to implement all the failover scenarios in Erlang than in C. So
far, here's a prototype of his: https://github.com/lemenkov/erlpmd
>
>
>
> When hacking it, I've noticed several things:
>
>
>
> 1. When we send ALIVE2_REQ and reply with ALIVE2_RESP, we establish a TCP
connection. Closing of which is a signal of node disconnection. This
approach does have a point, since we can use keep-alive and periodically
check that the node is still here on the TCP level. But next, some weird
thing follows:
>
>
>
> 2. When we send other control messages from a node connected to epmd, we
establish a new TCP connection, each time. Could use the main connection
instead. Was it a design decision or it's just a legacy thing?
>
>
>
> 3. The client (node) part of epmd seems to be all implemented in C and
sealed inside ERTS. However, it seems like this code could be implemented
inside the net_kernel module instead (or something similar).
>
>
>
>
>
> Why bother and switch to Erlang when everything is already written and
working? First of all, sometimes it doesn't really work in big clusters
(see my first link). And, secondly, using Erlang we can easily extend the
protocol. For example, add auto-discovery feature, which has been discussed
on the list a lot. Add an ability for a node to reconnect if its TCP
session has been terminated for some reason. Add lookups of nodes by prefix
(like, "give me all nodes that match mynode@*"). The list can be probably
extended further.
>
>
>
>
>
> Do you think such a thing (with full backwards compatibility, of course)
could go upstream? Also, a question for epmd maintainers: is it going to
change at all, or the protocol is considered to be full enough for its
purposes?
>
>
>
> --
> Best regards,
> Dmitry Demeshchuk
>
>
> _______________________________________________
> erlang-questions mailing list
> erlang-questions@REDACTED
> http://erlang.org/mailman/listinfo/erlang-questions
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://erlang.org/pipermail/erlang-questions/attachments/20121108/7db1ab55/attachment.htm>


More information about the erlang-questions mailing list